[llvm] [LLVM][NVPTX]Add BF16 vector instruction and fix lowering rules (PR #69415)

[Artem Belevich via llvm-commits]

================
@@ -1623,8 +1623,10 @@ SDValue SelectionDAGLegalize::ExpandFCOPYSIGN(SDNode *Node) const {
                                 SignMask);
 
   // If FABS is legal transform FCOPYSIGN(x, y) => sign(x) ? -FABS(x) : FABS(X)
+  // We don't do it in bf16 since the other path has less number of instructions
----------------
Artem-B wrote:

I agree that it may be better for NVPTX. I'm saying that *other* targets may or may not benefit from it the same way.

I think the decision should be delegated to TLI.

https://github.com/llvm/llvm-project/pull/69415